Errors made during V(D)J recombination, the process that assembles antigen receptor genes, can lead to chromosomal translocations and the development of human malignancies, particulariy childhood leukemias. One important cause of such translocations is the improper targeting of the RAG1 and RAG2 proteins, which constitute the central components of the V(D)J recombination machinery. In the first phase of V(D)J recombination, DNA substrate recognition and cleavage take place in highly organized nucleoprotein complexes whose integrity and specificity are determined by f^GI and RAG2, probably in conjunction with the DNA bending protein HMGB1. Little is known about the structure of these complexes, the conformational changes that occur during DNA binding, or the regulation of complex formation in vivo. Using novel biochemical, biophysical, and in vivo methodologies developed for the study of the RAG proteins during the previous funding period, three major aims will be pursued: 1) To understand the structure and dynamics of RAG-DNA complexes and determine how structural changes relate to RAG function; work under this aim will result in refined three-dimensional models of the DNA in fRAG-DNA complexes as well as the characterization of the pathway leading to complex formation. 2) To identify the role of HMGB1 in DNA binding and cleavage by RAG; work under this aim will take advantage of fluorescently-labeled HMGB1 and will reveal the stoichiometry, positioning, and function of HMGB1 in RAG-DNA complexes, as well as its contribution to the targe DNA bends identified in the prior funding period. 3) To determine the patterns of RAG1 and RAG2 binding throughout the genome of mouse and human developing lymphocytes; work under this aim will lead to an understanding of the rules that govern RAG recruitment to recombination centers in antigen receptor loci and to the thousands of sites they occupy elsewhere in the genome. Structural information from aims 1 and 2 will be connected to function through the use of altered DNA substrates and mutant RAG and HI\^GB1 proteins, and will be used to formulate models of how DNA recognition occurs within the genome. These models will be tested through in vivo binding experiments in developing lymphocytes and in cell line models. These studies should provide insights into how the RAG proteins and DNA communicate with one another at the molecular level and how this interaction leads both to carefully orchestrated antigen receptor gene assembly and to mis-recognition errors that underlie genome instability.